Endogenous Money: What it is and Why it Matters Abstract Endogenous money is widespread in economic theory. The Post Keynesian contribution is identification of a causal link between bank lending and the money supply. Though driven by macroeconomic concerns, the Post Keynesian debate has reduced to a microeconomic debate over the role of financial intermediaries in the accommodation process. In the ISLM model endogenous money flattens the LM. This misses its substantive significance which is discrediting of monetarist money supply policy rules and monetarist critiques of central banking, its identification of the key role of credit, and its provision of a credit driven theory of the business cycle. Key words: Endogenous money, bank lending, credit, business cycle. JEL ref.: E3, E4, E5. Thomas I. Palley Public Policy Department, AFL-CIO 815 Sixteenth Street, N.W. Washington DC 20006 Revised August 2001 Final revisions November 2001
Transcript
Endogenous Money: What it is and Why it Matters
Abstract Endogenous money is widespread in economic theory. The Post Keynesian contribution is identification of a causal link between bank lending and the money supply. Though driven by macroeconomic concerns, the Post Keynesian debate has reduced to a microeconomic debate over the role of financial intermediaries in the accommodation process. In the ISLM model endogenous money flattens the LM. This misses its substantive significance which is discrediting of monetarist money supply policy rules and monetarist critiques of central banking, its identification of the key role of credit, and its provision of a credit driven theory of the business cycle. Key words: Endogenous money, bank lending, credit, business cycle. JEL ref.: E3, E4, E5.
Thomas I. Palley Public Policy Department, AFL-CIO
815 Sixteenth Street, N.W. Washington DC 20006
Revised August 2001
Final revisions November 2001
1
INTRODUCTION
Endogenous Money occupies a central place in Post Keynesian economics. Within the
Post Keynesian schema, the money supply is endogenously determined via the demand
for bank credit. For the last decade, the Post Keynesian approach to endogenous money
has become bogged down in a debate between what have become labelled the
"accommodationist" and "structuralist" approaches (Pollin, 1991). Accommodationists
argue that the money supply is exclusively credit driven, with the monetary authority
setting the interest rate and being forced to accommodate any increase in demand for
reserves caused by increased bank lending. Structuralists also maintain that the money
supply is influenced by the demand for credit and the reaction of the monetary
authorities, but they argue that the money supply also depends on the asset and liability
management practices of banks (Palley, 1987, 1994a). Even if the monetary authority
refused to accommodate any increase in the demand for reserves, banks would still be
able to partially accommodate an increase in loan demand through their own initiatives.
Paradoxically, though the initial impulse for developing a theory of endogenous
money came from macroeconomics, the Post Keynesian debate has become centered on
the microeconomic behavior of the banking firm and its response to changed credit
market conditions. This paper seeks to restore a macroeconomic focus to the debate by
exploring the question of why endogenous money matters.
2
The paper begins with an examination of different forms of money endogeneity. There
are many forms of money endogeneity, and conventional neo-classical monetary theory
already incorporates certain forms. Thereafter, the paper examines macroeconomic
representations of endogenous money, and argues that endogeneity of money matters for
both short run comparative static macroeconomics and longer run macro dynamics. The
critical significance of endogenous money lies in the non-neutrality of money, which
means that financial arrangements are central to real macroeconomic outcomes.
Acceptance of Post Keynesian endogenous money theory implies that fluctuations in
monetary aggregates can be driven endogenously. This has a number of implications.
First, observation of a correlation between macroeconomic failure and contraction of
monetary aggregates proves nothing about policy as cause, which challenges monetarist
claims that macroeconomic failures are largely due to poorly executed central bank
control of the money supply (Palley, 1993).1 Second, the endogeneity of money means
that attempts to control the economy through monetarist styled money supply rules and
targets are likely fail.2 This suggests that policy authorities should look to other means of
control. Interest rate policy is one instrument of control, but there may also be a place for
1. The monetarist claim that central banks are responsible for macroeconomic stability can be illustrated through the Fisher equation, MV = PY. According to monetarism the money supply (M) is exogenous, money demand (1/V) is stable, and movements in the money supply temporally precede movements in nominal income (PY). Putting the pieces together, this implies that central bank changes in the exogenous money supply are the principal cause of macroeconomic fluctuations. The importance of Friedman and Schwartz’s (1963a, 1963b) work on the monetary history of the U.S. and the Great Depression is that it aimed to empirically verify this claim.
2. The failure of monetarist policy rules is empirically confirmed by the episode of monetarist policy dominance in the U.K. and U.S. in the late 1970s and early 1980s. This was a period characterized by enormous interest rate and real output volatility. Goodhart’s law (attributable to Charles Goodhart), that any stable relationship between a monetary aggregate and economic activity will break down if the monetary authority tries to control that aggregate, is the cynic’s version of the Post Keynesian claim that it is impossible to control the money supply.
3
quantitative regulation. Third, endogenously driven fluctuations of the money supply play
an important role in the business cycle and can contribute to instability. Regrettably, the
dynamic implications of Post Keynesian monetary theory have remained relatively
undeveloped. So too have its policy prescriptions, and this has diminished the Post
Keynesian contribution to the debate over operational monetary policy.
COMPETING FORMS OF ENDOGENOUS MONEY
Post Keynesian monetary theory rejects the notion of exogenous money, and maintains
that both neo-Keynesian and classical macroeconomics assume that the money supply is
exogenous. This section argues that such a claim is overstated, and that there exist a
range of distinct approaches to endogenous money. Thus, the Post Keynesian innovation
is not the distinction between exogenous and endogenous money, but rather the
construction of endogenous money in terms of bank lending.
(a) Evolutionary (Mengerian) endogenous money.
Neo-classical competitive general equilibrium (GE) theory is frequently represented as
paradigmatic of the exogenous money approach. However, GE theory has its own theory
of endogenous money.3 Within the GE perspective, money endogeneity can be
approached in two different ways. The first concerns the selection of a money, the focus
here being money as a medium of exchange and money’s ability to reduce transaction
("shoe leather") costs associated with exchange. This approach begins with Menger's
(1892) examination of the origins of money whereby the adoption of a particular
commodity as money is explained by the characteristic of "saleability". A commodity
gets adopted as money because it economizes on resources used in effecting exchange,
thereby making agents better off. Niehans (1971) presents a general equilibrium model in 3. An anonymous referee noted that though money can be introduced into the GE model, doing so requires the introduction of frictions. Such frictions are a somewhat arbitrary procedure considered by the standards of pure theory of the Arrow - Debreu model of perfect markets. Moreover, these approaches only work for limited aspects of money - principally, the medium of exchange function.
4
this spirit in which different commodities have different transactions costs, and the
commodity with the lowest transactions costs is adopted as money. King and Plosser
(1986) present a similar model in which money lowers the cost of trade, thereby enabling
agents to enjoy the benefits of specialization in production. In effect, the gains from
specialization underwrite the costs of trade. The selection of a commodity as money
involves balancing the commodities ability to facilitate low cost transacting against the
opportunity cost of having the commodity tied up as money rather than being used for
consumption or production. Finally, Kiyotaki and Wright (1992) present a search
theoretic account of the adoption of money in which money lowers the costs of searching
out trading partners, and different commodities have different search characteristics. The
important feature of these competitive general equilibrium accounts of money is that a
commodity is endogenously selected as money on the basis of its ability to facilitate
exchange at low cost. What constitutes money is therefore endogenous.
Selgin and White (1987) extend this neo-classical choice theoretic approach to a
dynamic context, and show how money endogenously evolves from commodity money to
fiat money and inside money. The force behind this evolution is the profit motive, which
sets up incentives to economize on the amount of money sitting in idle stores. Money
evolves endogenously, and the adoption of commodity money is just the first step in an
on-going process of financial innovation. Thus, monetization of an economy is
endogenous, as is the changing form of money.
(b) Neo-classical quantitative endogenous money
The selection of the form of money is one source of endogeneity in the competitive
general equilibrium model. The determination of the quantity of money is a second
source of endogeneity. This can be illustrated by reference to a gold standard economy in
which the total stock of gold is given at any moment in time, but its allocation between
monetary and non-monetary uses is endogenously determined by balancing the
5
opportunity cost of having gold tied up as money against the interest rate that can be
earned by placing gold on deposit with a bank. This process of endogenously determining the allocation of the gold stock can be described by the following three equation system: - + - + + + (1) G = GN(i, y, 1/P) + GM(i, y, 1/P) [Allocation of gold stock] + + (2) GM(i, y) = GB [Banking sector equilibrium] - + (3) GB = kPL(i, y) [Bank demand for gold] where G = total stock of gold GN = non-monetary demand for gold GM = gold deposits with the banking system GB = banking systems demand for gold i = interest rate y = level of income P = general price level k = fiduciary ratio L = demand for money balances
ensures that the total gold stock is allocated between monetary and non-monetary uses.
Equation (2) ensures that the banking system's demand for gold equals the supply of gold
available to the banking system. Equation (3) determines the banking systems demand for
gold which is proportional to the demand for inside money. The non-monetary demand
for gold depends negatively on the interest rate, positively on income, and negatively on
the relative price of gold compared to commodities. The price of gold is fixed at unity.
Substituting (2) and (3) into (1) then yields a gold standard LM schedule given by (4) G = GN(i, y, 1/P) + kPL(i, y)
The gold standard exhibits "static endogeneity". Changes in variables affecting either the
level of income or interest rates will affect the allocation of gold between monetary and
non-monetary uses, thereby affecting the money supply.4 This allocational endogeneity
has parallels with Keynesian - circuit monetary theory which is examined below.5
4.This model is capable of explaining the famous Gibson paradox which links interest rates to the price level. This can be done by adding an IS schedule and a positively sloped marginal cost curve given by - +
6
In addition to exhibiting static monetary endogeneity, the gold standard exhibits
dynamic monetary endogeneity. This is because the supply of gold can respond
endogenously over time. Such dynamic endogeneity can be incorporated by adding the
following equation
(5) MC(dG) = 1/P
where MCG = real marginal cost of producing gold, and dG = within period addition to
the gold supply. Gold producers mine new gold until the marginal cost of production
equals the price of gold. This ability to "produce" new money has a parallel in
contemporary new classical real business cycle theory which is examined below.
In sum, the neo-classical general equilibrium paradigm is normally viewed as having
exogenous money. In fact, it has its own forms of endogenous money. First, what serves
as money is endogenously determined. Second, the allocation of the commodity serving
as money (e.g. gold) between monetary and non-monetary uses is endogenously
determined. Third, the quantity of the commodity serving as money is subject to
endogenous influence.
(c) Fusions of evolutionary and neo-classical quantitative endogeneity
Neo-classical New Monetary Economics (NME) can be interpreted as fusing
evolutionary and quantitative endogeneity.6 According to NME, existing monetary
arrangements are predicated on the artefact of legal restrictions regarding the medium of
(4') y = y(i, G) + (4") P = MC(y) Positive IS shocks now cause a rightward shift along the LM which raises interest rates and output. At the same time prices rise. 5. This gold standard model can also be rendered compatible with Post Keynesian credit driven endogenous money by allowing the money supply to be determined by bank lending. In this event, increases in loan demand would increase the supply of deposits which would increase banks' demand for monetary gold, thereby causing interest rates to rise. 6. The legal restrictions theory of money is developed by Bryant and Wallace (1980).
7
payment for taxes. Absent this legal privileging of government issued money,
competitive markets would generate pure private monetary arrangements under which
both the means of payment and its quantity would be endogenous. For instance, Black
(1970) and Fama (1980) consider the economics of a completely unregulated banking
system in which banks are a form of mutual fund providing deposits backed by holdings
of real assets. This leads to the notion of mutual fund money. The standard of value is
shared by all, but the means of payment is title to real claims, and bank settlements
involve transfers of titles between banks.
Such competitive unregulated banking systems resemble Wicksell’s (1898) pure
credit economy, and they seem to suffer from many of the same problems (Trautwein,
1997). The essential problem is that the price level can be explosive owing to unlimited
bank lending that expands the nominal money supply. Wicksell (1898) dealt with this by
recourse to the assumption of cash drains, but this re-equilibrating tendency is not based
on the properties of a pure credit economy since cash is not used. In NME models,
prevention of an exploding price level seems to require 100% reserve requirements
whereby all deposits are always fully backed by real assets, thereby precluding bank
expansion of credit. However, even this may not be enough since an asset bubble driven
by expectations could drive up the value of mutual fund money, thereby driving up the
price of output and reinforcing the bubble.
Hicks (1989) also developed a theory of endogenous money embedding evolutionary
and quantitative endogeneity. He envisages monetary arrangements as evolving toward
Wicksell’s pure credit economy, with non-interest bearing money ceasing to be a reserve
asset and all money becoming just a debt. His resolution of the excessive lending - price
indeterminacy problem rests on lack of sufficient trustworthy borrowers. This means that
there could even be a shortage of investment relative to savings - the exact opposite of the
Wicksell’s problem. Such a shortage could result in deflationary pressures, which in a
credit economy would generate instability through Fisher (1933) style debt deflations.
8
These speculations about the operation and stability of endogenous money pure
credit economies suggest the possibility of a convergence with the Post Keynesian
approach. However, there are significant impediments to such a convergence. First, Post
Keynesians (Davidson, 1978), emphasize the demand for liquid stable value assets, and
this feature is absent in pure credit economies which raises doubts about their analytic
relevance.7 Second, within Wicksell - NME pure credit economies the equilibrium real
interest rate is determined by the fundamentals of tastes, technologies, and endowments.
Amongst Post Keynesians the issue is more open. Accommodationists maintain that the
real interest rate is determined by central banks and behavioral conventions. Structuralists
see it as being influenced by both of these factors, and also impacted by structural factors
including the state of the real economy and the state of financial intermediary balance
sheets. Third, in Wicksell - NME pure credit economies the level and growth of
equilibrium output are again determined by fundamentals (tastes, technologies, and
endowments), whereas Post Keynesians see an independent place for the level and
growth of aggregate demand - both of which may be impacted by monetary
arrangements.
(d) Central bank endogeneity
The behavior of central banks in response to economic developments generates a
completely different form of endogeneity. Within macro models this form of endogeneity
is captured in the central bank's policy response function. There is a substantial literature
on this subject, initiated by Poole (1970), examining the question of optimal monetary
stabilization policy.8 The core question is whether the monetary authority should target
interest rates, the money supply, or some combination of the two. Poole’s (1970) model 7. Palley (2000) examines the implications of the e-money revolution which may open the way to a financial system that more closely resembles a pure credit economy. In such an economy there stands to be greatly heightened financial instability owing to the increased elasticity of private credit money production, combined with periodic liquidity crises generated by runs out of bank e-mutual fund money into central bank money. 8. This literature is surveyed in Friedman (1990).
9
has the monetary authority directly controlling the money supply. In more sophisticated
models (Friedman, 1974) the monetary authority lacks direct control over the money
supply, and instead influences it indirectly through control over the monetary base.
Central bank reactions fuse concerns of "endogeneity" with those of "controllability".
The money supply may be endogenous through the central banks response function and
perfectly controllable. Alternatively, it may be endogenous and imperfectly controllable.
In the former case, endogeneity is exclusively the product of the central bank's response
function. In the latter case, endogeneity is partly autonomous, reflecting money supply
disturbances generated from within the economy and outside of central bank control.
The Post Keynesian theory of credit driven endogeneity is not intended to be a theory
of central bank endogeneity. Yet paradoxically, in accommodationist accounts it is the
central bank’s response function (i.e. the willingness to accommodate increased demand
for reserves at an unchanged interest rate) that is the mechanism enabling the banking
system to support expanded bank lending. This contrasts with the structuralist account in
which the key mechanism is banks’ asset and liability management activities, and the
central bank’s response function serves as an additional enabling mechanism.9
Why does the central bank accommodate monetary disturbances? Neo-Keynesian
optimal stabilization theory maintains that it does so voluntarily for purposes of optimally
stabilizing the economy. Post Keynesian accommodationists argue that accommodation is
“institutionally” forced by the existence of loan commitments. However, the existence of
such commitments does not mean that they are necessarily exercised. Though banks
make promises to lend, the rate at which they lend remains variable and market
determined. Interest rates can therefore rise and induce agents not to exercise their loan
commitments, so that higher interest rates represents a means of de facto extinguishing
loan commitments. Consequently, lines of credit do not imply a horizontal money supply.
9. The significance of the central bank’s reaction function for the process of accommodation is explored in Palley (199a).
10
Another explanation of central bank accommodation can be labelled "political
endogeneity". Here, the reasoning is that central banks must accommodate the increase in
deposits caused by increased bank lending, or else the banking system will be rendered
insolvent and collapse as banks will be unable to obtain the needed reserves. This claim
introduces issues of expectations about policy, time consistency, and credibility. The
extent to which central banks have no choice will depend on the policy regime, and on
the expectations of financial institutions (FIs). If FIs believe that the existing regime is
lax, then they will structure their lending behaviors and balance sheets on the basis that
the central bank will accommodate. Under these conditions, refusal to accommodate
could lead to collapse. However, if FIs believe the regime is tight, they will structure their
activities accordingly and maintain access to liquidity. Consequently, the system will be
less vulnerable to non-accommodation. In sum, central banks do have a choice, but that
choice depends on signals sent about the policy regime and the degree to which these
signals are credible.
(e) Fiscal endogeneity
Another form of endogeneity can be labelled "fiscal endogeneity". It emphasizes the
government budget constraint and the fact that government deficits/surpluses have
implications for asset stocks. These features are emphasized by Christ (1968) and Blinder
and Solow (1973). The government budget constraint is given by
(6) D = G - T + iB = dB + dM
where D = deficit or surplus, G = government spending, T = tax revenues, i = interest
rate, B = national debt, dB = change in national debt, and dM = change in monetary base.
To the extent that deficits are money financed, they increase the money supply.
Kaldor (1970) emphasized this fiscal endogeneity in his initial critique of
monetarism.10 Monetarists argue that money is all that matters, and that fiscal policy is
10. Kaldor (1980, footnote p.307) distanced himself from the idea that the government budget constraint was important for explaining money supply endogeneity.
11
irrelevant. Kaldor argued that the apparent statistical significance of money reflected the
operation of the government budget constraint, with increased money supplies being the
result of money financed government spending. Tobin (1970) also emphasized the role of
the government budget constraint in rejecting monetarist claims about the importance of
money predicated on leads and lags in money growth.
Consideration of the link between the government budget constraint and the money
supply also compels consideration of central bank governance. To the extent that central
banks are independent, governments lack direct control over the money supply and must
in the first instance finance deficits exclusively with bonds. Thereafter, the money supply
changes only if the central bank decides to accommodate the deficit, but this is a matter
of central bank policy.
(f) The money multiplier and portfolio endogeneity
Conventional textbook accounts of money supply determination rest on money
multiplier theory. The money multiplier approach embodies "portfolio endogeneity"
whereby the money supply depends on agents’ portfolio preferences for money relative to
other assets. The magnitude of the multiplier depends on the mix of assets held, and the
multiplier is endogenous to the extent that asset demands vary in response to interest rate
and income changes. Thus, higher interest rates induce agents to economize on liquidity,
causing both the money multiplier and the money supply increase.
Changes in the money multiplier arise from changes in the composition of portfolios.
In this regard there are two sets of portfolios -- those of the non-bank public and those of
banks. The non-bank public consists of households and non-bank firms, both of which
demand cash, checkable deposits, and time deposits. Variations in these demands are the
traditional focus of money multiplier analysis. Banks also demand liquidity in the form
of excess reserves, so that changes in bank preferences also affect the money multiplier.
These portfolio composition effects are an integral part of the Post Keynesian
structuralist approach to the money supply. They provide a mechanism whereby banks
12
are able to generate additional liquidity to support deposits created by increased bank
lending. Bank asset and liability management activities are central. Banks vary their own
portfolio demands (asset management), and induce variations in the portfolio demands of
the non-bank public (liability management). Together, these variations allow banks to
accommodate increased lending independent of the monetary authority’s stance.
(g) Post Keynesian endogenous credit money
The neo-Keynesian money multiplier emphasizes changes in the composition of
agents' portfolios, but makes no mention of the lending activities of banks. Contrastingly,
emphasis on bank lending is the distinguishing hallmark of the Post Keynesian approach
to endogenous money. Post Keynesians maintain that bank lending creates deposits once
loans are credited to borrowers’ accounts, and this shifts attention toward the asset side of
bank balance sheets. This credit driven approach to the money supply is captured in
Coghlan's (1978) reformulation of the money multiplier in terms of a loan multiplier.11
However, having identified the significance of bank lending for the money supply,
Post Keynesians divide over the process by which deposits created by lending are
supported. Accommodationists assert that bank asset and liability management is
irrelevant, and the central bank automatically provides whatever liquidity is provided
(Moore, 1988).12 Structuralists maintain that bank asset and liability management is
important, that there is no necessary requirement for the central bank to provide liquidity,
and the process by which the central bank injects liquidity into the system also matters
11. A simplified banking sector balance sheet is given by Loans (L) + Excess Reserves (eD) + Required Reserves (kD) = Deposits. Collecting terms yields D = L/[1 - e - k]. The quantity of deposits is therefore a multiple of the level of bank lending. 12. In recent years there seems to have been a blurring of this distinction with accommodationists claiming that bank asset and liability management is part of their understanding of the accommodation process. Yet, as Tobin (1963) makes clear, agents have limited demands for the liabilities offered by banks. This means that banks will have to pay agents more to induce them to increase their holdings, and loan rates will therefore rise as banks asset and liability management to accommodate increased lending.
13
(Palley, 1996b). Though not the original intent of the theory of endogenous money, these
differences have contributed to making the microeconomics of the banking firm the focus
intermediary supply-side endogeneity, monetary circuit endogeneity, and open economy
endogeneity. This section addresses the macroeconomic implications.
Post Keynesian endogenous money and ISLM old Keynesianism
The ISLM model has long served as the canonical model of macroeconomic analysis.
Classical monetarism is identified most closely with the doctrine of exogenous money. In an ISLM framework, it is described by the following equations: - + (8) y = D(i, y) [IS schedule] (9) Ls = ky k > 0 [LM schedule]
where y = income, D = aggregate demand, i = interest rate, and Ls = real money supply.
this system of equations in [y, i] space. The IS is negatively sloped, while the LM is
vertical. The important feature is that the money supply is fixed, and there is no
substitutability regarding money demand. Consequently, there is a binding monetary
constraint on economic activity which cannot be escaped.13
Replacing equation (9) with Keynes' General Theory construction of money demand
yields the following system: - + (8) y = D(i, y) [IS schedule] (9.1) Ls = L1(y) + L2(i) = Ld(i, y) [LM schedule] 13. Price deflation which expands the real money supply provides the one avenue for the system to endogenously escape the monetary constraint. However, this avenue is predicated on the absence of negative debt effects. This issue is discussed subsequently.
17
Figure 2 shows this system of equations in [y, i] space. The only change from Figure 1 is
that the LM schedule is positively sloped. The significance of this is that the financial
system can accommodate an increase in the demand for money caused by an increase in
the level of economic activity. Expansionary shifts of the IS are no longer fully crowded-
out. In effect, speculative balances serve as a form of quasi-endogenous money. When
transactions needs rise, rising interest rates induce wealth holders to release part of their
speculative balances to meet these needs. Note exactly the same outcome holds if money
demand is described in terms of a twin circuits approach, as described by equation (7.1).
If the money supply is allowed to respond endogenously to the level of interest rates
as per the money multiplier story, the benchmark model becomes - + (8) y = D(i, y) [IS schedule] +
(9.2) Ls(i) = ky [LM schedule]
Once again the LM schedule is positively sloped in [y, i] space, and the graphical
representation of the model is exactly as in Figure 2. The qualitative properties of this
model are identical to those of the standard model. This is because an increase in the
money supply is isomorphic with a decrease in money demand: having the money supply
increase in response to rising interest rates has the same qualitative effects as having
money demand decrease. If (9.1) and (9.2) are combined, the benchmark model becomes - + (8) y = D(i, y) [IS schedule] + - + (9.3) Ls(i) = Ld(i, y) [LM schedule]
The money supply function and Keynesian money demand function both serve to flatten
the LM schedule. Were this the sum contribution of endogenous money, it would add
little to the analytics of the benchmark neo-Keynesian macro model.
18
Thus far, endogenous money has been characterized as flattening the LM schedule,
but the IS and LM schedules have remained independent of each other. In a monetary
economy transactions in the real sector always have a financial counterpart, which means
that the IS and LM schedules will tend to co-move. Davidson (1965) presents an ISLM
model that has this feature. The important innovation is that money demand is now scaled
by consumption and investment spending rather than income, reflecting a - + + (9.4) Ls = Ld(i, C, I) [LM schedule] - + - (10) C = C(i, y, A1) [Consumption function] - + (11) I = I(i, A2) [Investment function]
where C = consumption spending, I = investment spending, A1 = shock to consumption,
and A2 = shock to investment. The IS and LM schedules have the conventional slopes in
[y, i] space, but shocks to consumption and investment spending which shift the IS
schedule also impact money demand and shift the LM schedule. Positive expenditure
shocks therefore shift the IS right, and the LM left. Interest rates rise unambiguously, but
the effect on income is ambiguous.
The significant contribution of Davidson's model is that it structurally links
movements in the IS and LM schedules. In doing so, it addresses Robertson's (1940)
loanable funds critique of the liquidity preference theory of interest rate determination.
Now, decreases in thrift and increases in the marginal efficiency of investment both have
an immediate positive effect on interest rates. However, Davidson's model also increases
the extent of crowding-out, and increases in consumption and investment spending can
lower output if the induced increase in money demand is sufficiently large.
Analytically, Davidson’s finance motive construction represents an early middle
ground between the ISLM neo-Keynesian approach to money which emphasized a
“money to hold” perspective and the new credit approach which emphasizes a “money to
19
spend” perspective.14 This bridge can be further improved by making the money supply
endogenous, thereby reflecting the impact on the money supply of bank lending to
finance spending. This can be done by having the money supply depend on consumption
and investment, and the model becomes: - + (8) y = D(i, y) [IS schedule] + + + - + + (9.5) Ls(i, C, I) = Ld(i, C, I) [LM schedule] - + - (10) C = C(i, y, A1) [Consumption function] - + (11) I = I(i, A2) [Investment function]
The slope of the IS schedule is unchanged, but the LM is flatter. Moreover, the
comparative statics are also altered. In the Davidson model, positive shocks to
consumption and investment spending increase money demand, and cause the LM to shift
upward. Now, such shocks also cause the money supply to increase. Thus, the upward
shift of the LM schedule is diminished.15
The macro model suggests that the LM schedule could even shift down if the increase
in the money supply were sufficiently large. However, this would be inconsistent with the
microeconomic account of endogenous money. According to that account, increased bank
lending to finance consumption and investment spending leads to the creation of deposits.
The banking system needs liquidity to back these deposits, and must engage in asset and
liability management to accomplish this. These measures rely on higher interest rates,
which means that the LM cannot shift down. Thus, the microeconomic logic of
endogenous money constrains the macroeconomic possibilities. 14. The distinction between “money to hold” and “money to spend” is drawn from Trautwein (2000). 15. This augmented version of Davidson’s (1965) model has resemblances with Bernanke and Blinder’s (1988) new Keynesian model of the credit transmission framework. The endogeneity of the money supply reflects the effects of lending on the money suppl, and lending is also assumed to positively impact the IS. In the Bernanke and Blinder model the effects of lending are restricted to the credit market.
20
The link between Post Keynesian endogenous money and new Keynesian economics
Attempts to place endogenous money within the old Keynesian ISLM framework
involve tinkering with the specification of the LM schedule. However, the heart of
endogenous money - at least for Post Keynesians - involves recognizing the link with
bank lending. This suggests explicitly modeling the credit market.
Bernanke and Blinder (1988) do this, and provide a tractable macro model in which
the IS schedule is replaced by a CC schedule (commodities and credit) which embeds
equilibrium in both the goods and credit markets. This model has become labeled as
“new Keynesian,” but analytically it parallels Post Keynesian structuralist models as
developed by Palley (1987, 1994a).16 The big representational difference is that Bernanke
and Blinder rely on the “black box” of the money and loan multipliers to endogenize the
money supply, whereas structuralist models open the black box and explain it in terms of
the profit maximizing asset and liability management actions of banks. The financial side of the Bernanke - Blinder (1988) model is given by - + + (12.a) D(iB, y) = mD(iB)R - + + + + - + (12.b) L(iL, iB, y) = mL(iL, iB)D(iB, y)[1 - k]
where D(.) = demand for bank deposits, L(.) = demand for loans, mD = money multiplier,
mL = loan multiplier, R = monetary base, iB = bond interest rate, iL = loan interest rate,
and k = reserve requirement ratio. Equation (12.a) is the money market equilibrium
condition, while (12.b) is the credit market equilibrium condition. Substituting (12.a) into
(12.b) yields
16. The analytic equivalence of the models can be seen through the number of market equilibrium conditions. Both have a goods market, a money market, a credit market, and a bond market. Walras’ law allows for dropping one market equilibrium condition, and both drop the bond market. The Bernanke and Blinder (1988) commodities - credit (CC) schedule is obtained by substituting the credit market equilibrium condition into the goods market equilibrium condition. Palley (1987) also allows for credit market shocks to directly impact the goods market in the spirit of Davidson (1965).
21
(12.c) L(.) = mL(.)mD(.)R[1 - k]
The amount of bank lending is therefore determined by multiplicative interaction of the
loan and money multipliers. Higher interest rates raise both multipliers, thereby
increasing the quantity of bank liabilities and assets. Increased loan demand is
accommodated by a rise in interest rates, which increases the loan and money multipliers,
thereby enabling financial intermediaries to meet higher loan demand.
This process is identical to that described in micro-founded Post Keynesian
structuralist models. In effect, the money multiplier (mD) performs the role of bank
“liability” management, while the loan multiplier (mL) performs the role of bank “asset”
management. The benefit of a micro-founded approach is that it explicitly accounts for
these multiplier responses in terms of maximizing behavior by financial intermediaries.17
A rise in loan demand drives up loan market interest rates, inducing financial firms to
shift out of bonds into loans (asset management), which drives up bond rates. Firms also
raise rates paid on deposits to attract funds (liability management). These adjustments to
the asset and liability positions of financial firms are connected by first order conditions
requiring that the marginal cost of funds equal the marginal return to funds, that the
marginal cost of different sources of funds be equal, and that the marginal return to
different applications of funds also be equal.18 Profit maximizing behavior drives
financial firms to equalize marginal costs and returns across different financial markets. It 17. Trautwein (2000) speculates whether it is banks or financial intermediaries that matter for endogenous money, and comes to the conclusion that it is financial intermediaries - - “Banks should rather serve as a metaphor for all financial institutions that have the power to create assets of a high degree of liquidity (liabilities on their balance sheets. (P.181)”. This is a conclusion that I strongly agree with. The important analytic insight of endogenous money is it reveals how financial intermediaries can circumvent financial constraints on economic activity imposed by monetary authorities. With the decline of traditional banking it is time to move beyond endogenous money and start talking of endogenous finance (Palley, 1996c). 18. In a sense, banks and financial firms are multi-input - multi-output firms. On the input side they can draw funds from different sources - demand deposits, time deposits, commercial paper issues, etc. On the output side they can allocate their funds through loans to different types of borrower or they can hold government bonds, etc.
22
is this behavior that enables the financial sector to accommodate increases in credit
demand, explains the movement of loan and deposit multipliers, and explains why the
inside money supply responds positively to loan demand shocks. The bottom line is that
the Bernanke - Blinder (1988) new Keynesian construction of the money - credit nexus is
the same as the Post Keynesian structuralist perspective.19
Implications for dynamic macroeconomics
Viewed from the vantage of static old Keynesian macroeconomics, the debate over
endogenous money has a tendency to reduce to a debate over the slope of the LM.20
However, endogenous money also has implications for dynamic macroeconomics, and
this stands to be a fruitful area for future research.
Recall the ISLM construction of classical monetarism given by equations (8) and
(9). Under that construction the only way for real economic activity to escape the
monetary constraint is through price deflation. This raises the real money supply and
shifts the LM schedule right. However, the view that price deflation can loosen the
financial constraint is predicated on exogenous money and the absence of negative debt
effects. Once money is viewed as endogenous and created through the process of lending
- as is the case for the Post Keynesian approach - then deflation can lower economic
activity through its impact on debt burdens. Endogenous money, with its concomitant
recognition of debt, therefore provides an answer to the long standing debate in
macroeconomics as to whether market economies are self-equilibrating and Keynesian
unemployment is just a special case predicated on downward nominal wage rigidity.
19. New Keynesian microeconomic models of credit rationing (for instance Stiglitz and Weiss, 1981) appear perfectly compatible with the Post Keynesian perspective on endogenous money. Within existing Post Keynesian models bank lending is limited by the demand for loans at the market equilibrium interest rate. Given imperfect information, lending could be further restricted by banks’ willingness to lend for fear of defaults, and loan rates would also be altered so as to reflect costs associated with default. 20. This renders the issue of endogenous money reminiscent of the earlier debate between neo-Keynesians and monetarists (Tobin, 1974).
23
For Post Keynesians, endogenous money inevitably leads to the Fisher debt effect
(Fisher, 1933) which explains why price and nominal wage reductions cannot remedy a
deficiency of aggregate demand. There are several reasons why price deflation may
reduce output and employment. First, price deflation makes creditors wealthier at the
expense of debtors by raising the real value of nominal debts and debt service burdens,
and this reduces aggregate demand if creditors have a lower propensity to consume than
price expectations lower investment by giving agents an incentive to switch into money
away from real capital (Tobin, 1975). Third, deflation may also raise real interest rates
owing to the existence of nominal interest rates floors (Krugman, 2000), and owing to
increased risk of bankruptcy which raises credit risk (Palley, 2000). Fourth, in a world
where production takes time and firms incur costs before they realize revenues, deflation
may make it impossible for firms to recover their costs so that they reduce production
(Palley, 1997b). Fifth, deflation may push highly leveraged firms into bankruptcy. Not
only does this lower output immediately, it also destroys borrower - lender information
which restricts productive lending (Bernanke, 1983; Stiglitz, 1992; Calomiris, 1993).
The Post Keynesian construction of endogenous money helps explain how debt gets
into the economic system. The theory of endogenous money also contributes to providing
a financial theory of the business cycle. The Fisher debt effect is a “stock” effect. When it
comes to explaining the business cycle, it is changing "flows" of inside debt that matter.
These flows matter for the determination of AD, and it is here that the theory of
endogenous money makes an original contribution to macroeconomics. Palley (1994b)
presents a credit driven model of the business cycle in which debt is a two edged sword.
Increases in debt initially raise aggregate demand, but subsequent debt service payments
lower aggregate demand. A business cycle emerges because loan expenditures raise
income, which in turn generates fresh borrowing that further stimulates the economy.
However, debt service payments are contractionary and eventually come to dominate the
24
expansionary impact of new lending flows. This causes the cycle to turn, triggering a
period of contraction in which income falls and borrowings are repaid.21
This debt driven model of the business cycle can be expanded to include the effects of
endogenous money (Palley, 1997b). The key extension is the recognition that bank
lending creates money. The newly created money balances circulate as part of the flow of
money income and raise aggregate demand as long as the loans remain in circulation. The
fact that bank lending creates new money distinguishes it from bond market lending
(loanable funds lending). The latter involves a transfer of money balances from lender to
borrower, whereas the former involves the creation of new money balances. For this
reason, bank lending is more expansionary than bond market lending.
The above Post Keynesian models of the business cycle can be compared to new
Keynesian credit-driven business cycle models as developed by Kiyotaki and Moore
(1997) and Bernanke, Gertler, and Gilchrist (1996). There are similarities and
differences. Within new Keynesian models collateral requirements play a critical role,
protecting lenders against asymmetric information and moral hazard. This re-focuses the
monetary transmission mechanism away from the traditional cost of capital channel to
asset prices and collateral values. Changes in interest rates or productivity shocks both
impact asset prices and collateral values, which then affect firms’ abilities to borrow to
finance production.
In these new Keynesian models credit impacts economic activity through the supply-
side.22 Output is supply determined, and credit matters because it allows firms to finance
increased supply. In the Kiyotaki and Moore model lower asset prices tighten credit
constraints, which lowers capital investment and next period production, which then 21. The model can be given Minskyian (Minsky, 1982) features by having the allowable debt to income ratio fluctuate with the cycle. Thus, lenders raise debt ceilings when income is growing, and they lower them when income is falling. This lengthens the cycle and increases its amplitude. 22. This feature of new Keynesianism has led Davidson (1992) to question whether Keynes would have been a new Keynesian.
25
tightens next period’s credit constraint. When placed in a context with capital
depreciation, this can generate cycles. Borrowing to finance investment raises debt
burdens which restricts ability to invest, and if depreciation outweighs gross investment,
the capital stock falls. This then lowers next period production and borrowing capacity,
thereby triggering a period of contraction.
The similarity between Post Keynesian and new Keynesian models concerns the two
edged nature of debt. Borrowing finances economic expansion, but it also generates debt
burdens that contract activity. The big difference between the two approaches concerns
output determination. In new Keynesian models output is supply-driven. In Post
Keynesian models it is demand driven. This difference has implications for how debt
burdens cause cyclical expansions to end. In the Kiyotaki - Moore (1997) model capital
depreciation is critical, and the cycle peaks when depreciation comes to outweigh the
debt constrained level of investment. In Palley (1994, 1997) it is differences in the
marginal propensity to consume between creditors and debtors. Debt burdens are
contractionary because they transfer income from debtors to creditors, and the expansion
ends because the growing transfer of income outweighs the expansionary impact of new
borrowing.
Another difference is the new Keynesian emphasis on the asset price - collateral -
investment channel. However, this channel can be readily and sensibly incorporated in
the Post Keynesian framework. Thus, let investment spending be the minimum of desired
spending and unused borrowing plus free cash flow
(13) It = Min[ I*t, D*
t - Dt-1 + FCt]
where It = actual investment spending, I*t = desired investment spending, D*
t = firms’
debt ceiling, Dt-1 = firms’ last period debt, and FCt = free cash flow. Free cash flow is
given by
(14) FCt = yt-1 - iDt-1
Corporate debt ceilings can then be specified as multiple of business collateral
26
(15) D*t = aWt a > 0
where Wt = value of firms’ collateral. Lastly, the value of collateral can be made a
function of lagged, current, or expected output. Given this, the new Keynesian financial
accelerator can be readily incorporated in the Post Keynesian multiplier-accelerator
model of credit driven business cycles. If asset prices are a function of expected output,
this allows for Minsky (1982) style asset bubbles and manias that have real effects.
Finally, the theory of endogenous money may also prove relevant for the theory of
economic growth - though this is a conjecture. Keynesian growth theory emphasizes the
interaction of demand growth and supply growth, with the rate of demand growth
exercising an independent influence on the rate of growth of supply (Palley, 1996a). If
the rate of growth of demand depends on the rate of growth of lending, then credit and
endogenous money become relevant to the theory of economic growth. Faster loan
growth leads to faster demand growth, which leads to greater investment and faster
output growth. It is the endogeneity of money - the means of payment - that enables
demand to lead output. Not only does this provide an incentive to invest, but it also
avoids the damaging effects of deflation (discussed earlier) which would ensue if
productivity grew but nominal demand were fixed.
This suggested role of endogenous money in the growth process contrasts with
existing constructions of the role of money and finance in growth. It is different from
Tobin’s (1965) neo-classical monetary growth models which emphasizes the impact of
money on inflation, and the incentive inflation generates for portfolio substitutions
toward capital. It also differs from Schumpeter’s (1911) construction of finance and
growth which emphasizes the allocative role of financial intermediaries in getting
resources to entrepreneurs.
CONCLUSION
This paper has catalogued the many sources of endogenous money. The Post
Keynesian approach focuses on lending by financial intermediaries, which has some
27
similarities with neo-classical constructions of pure credit economies. However, in those
models money and credit exhibit long run neutrality, whereas in Post Keynesian models
money and credit are non-neutral.23
The Post Keynesian theory of endogenous money has both positive and
normative implications. It describes the critical role of lending in the creation of money,
and it shows how increased lending is supported by the portfolio choices of households
and the asset and liability management decisions of financial intermediaries. This is a
position that has been adopted by new Keynesians. It also elevates the role of credit in the
macroeconomic process. Credit is central to Post Keynesian economics, whereas it is
strikingly absent (as can be seen from inspection of standard macroeconomics textbooks)
in both neo-Keynesian and new classical macroeconomics. The presence of debt in a
monetary economy challenges claims about the capacity of price adjustment to restore
full employment equilibrium, which refutes charges that Keynesian unemployment is a
special case predicated on downward nominal wage rigidity. Lastly, the Post Keynesian
approach to endogenous money can be extended to the business cycle, and it illuminates
both the critical role of debt in the cycle and why the money supply moves pro-cyclically.
New Keynesian models of the business cycle also emphasize debt, but they have output
being supply- rather than demand-determined. These new Keynesian models also lack an
explicit role for money.
At a normative level, the Post Keynesian approach challenges monetarist claims that
macroeconomic fluctuations are the product of misguided monetary control by central
banks. It also challenges the efficacy of monetarist prescriptions recommending targeting
of monetary aggregates. Instead, it recommends interest rate policy and regulatory
controls - such as asset based reserve requirements - that automatically restrict the ability
of the financial sector to expand lending. Finally, rather than being a source of
23. Trautwein (2000) also emphasizes the long run neutrality of credit as a critical dividing line in his survey the economics of credit.
28
inefficiency, in a credit money economy downward nominal wage rigidity is a source of
economic stability that guards against deflation.
29
References Bernanke, B., “Non-Monetary Effects of the Financial Crisis in the Propagation of the Great Depression,” American Economic review, 73 (1983), 257 - 76. Bernanke, B., and Blinder, A., “Credit, Money, and Aggregate Demand,” American Economic Review, 78 (May 1988), 435 - 39. Bernanke, B., Gertler, M., and Gilchrist, S., “The Financial Accelerator and the Flight to Quality,” Review of Economics and Statistics, 78 (1996), 1 - 15. Black, F., “Banking and Interest Rates in a World Without Money. The Effects of Uncontrolled banking,” Journal of Bank Research, 1 (1970), 9 - 20. Blinder, A.S., and R.M.Solow, "Does Fiscal Policy Matter?" Journal of Public Economics, (November 1973), 319-37. Bryant, J., and Wallace, N., “A Suggestion for Further Simplifying the Theory of Money,” Staff Report 62, Federal Reserve Bank of Minneapolis, Minneapolis: MN., 1980. Calomiris, C., “Financial Factors in the Great Depression,” Journal of Economic Perspectives, 7 (1993), 61 - 85. Caskey, J., and S.Fazzari, "Aggregate Demand Contractions with Nominal Debt Commitments: Is Wage Flexibility Stabilizing?" Economic Inquiry, 25 (October, 1987), 583-97. Christ, C.F., "A Simple Macroeconomic Model with a Government Budget Restraint," Journal of Political Economy, 76 (1968), 53 - 67. Coghlan, R., "A New View of Money," Lloyds Bank Review, 129 (July 1978), 12 - 28. Davidson, P., “would Keynes be a New Keynesian?” Eastern Economic Journal, 18 (1992) ---------------, Money in the Real World, New York: John Willey and Sons, 1978. ----------------, "Keynes' Finance Motive," Oxford Economic Papers, 17 (1965). Fama, E., “Banking in the Theory of Finance,” Journal of Monetary Economics, 6 (1980), 39 - 57. Fisher, I., "The Debt-Deflation Theory of Great Depressions," Econometrica, 1 (October 1933), 337-57.
30
Fleming, J.M., “Domestic Financial Policy under Fixed and Floating Exchange Rates,” IMF Staff Papers, 9 (1962), 369 - 79. Friedman, B., "Targets, Instruments, and Indicators of Monetary Policy," Brookings Papers on Economic Activity, 1977-2 (1977), 293 - 335. ------------, "Targets and Instruments of Monetary Policy," in B. Friedman and F. Hahn (eds.), Handbook of Monetary Economics, Vol. 2, New York: North Holland, 1990. Friedman, M., and Schwartz, A., “Money and Business Cycles,” Review of Economics and Statistics, supplement February 1963a, 32 - 64. --------------------------------------, A Monetary History of the United States, 1867 - 1960, Princeton, NJ: Princeton University Press, 1963b. Hicks, J., A Market Theory of Money, Oxford: Clarendon Press, 1989. Kaldor, N., "The New Monetarism," Lloyds Bank Review, 97 (July 1970), 1 - 17. -------------, “Monetarism and U.K. Monetary Policy,” Journal of Economic Research, 4 (1980), 293 - 318. Keynes, J.M., A Treatise on Money, London: Macmillan, 1930. ------------, The General Theory of Employment, Interest and Money, London: MacMillan, 1936. King, R.G., and Plosser, C.I., "Money, Credit and Prices in a Real Business Cycle," American Economic Review, 74 (1984), 363-80. -----------------------------, "Money as the Mechanism of Exchange," Journal of Monetary Economics, 17 (1986), 93 - 115. Kiyoyaki, N., and Moore, J., “Credit Cycles,” Journal of Political Economy, 105 (1997), 211 - 48. Kiyotaki, N., and Wright, R., "Acceptability, Means of payment, and Media of Exchange," Quarterly Review of the Federal Reserve of Minnesota, Summer 1992, 18 - 21. Krugman, P., “It’s Baaack! Japan’s Slump and the Return of the Liquidity Trap,” Brookings Papers on Economic Activity, 1998, 137 - 205. Menger, K., "On the Origin of Money," Economic Journal, 2 (1892), 239 - 55. Minsky, H.P., Can “It” Happen Again?, Armonk, NY: M.E. Sharpe, 1982.
31
Moore, B.J., Horizontalists and Verticalists: The Macroeconomics of Credit Money, Cambridge University Press: Cambridge, 1988. Mundell, R., International Economics, New York: Macmillan, 1968. Niehans, J., "Money and Barter in General Equilibrium with Transactions Costs," American Economic Review, 61 (1971), 773 - 83. Palley, T.I., “The e-Money Revolution: Challenges and Implications for Monetary Policy,” Journal of Post Keynesian Economics, forthcoming. -------------, “The Case for Equilibrium Low Inflation: Some Financial Market Considerations with Special Attention to the Problem of Japan,” Eastern Economic Journal, 26 (Summer 2000), 277 - 97. -------------, “General Disequilibrium Analysis with Inside Debt,” Journal of Macroeconomics, 21 (Fall 1999), 785 - 804. ------------, "The Twin Circuits: Aggregate Demand, Liquidity Preference, and the Expenditure Multiplier," Review of Radical Political Economics, 30 (September 1998), 91 - 100. ------------, “Expectations, the Production Period, and Keynes’ Aggregate Supply Schedule,” Manchester School, LXV (June 1997), 295 - 309. ------------, "Endogenous Money and the Business Cycle," Journal of Economics, 65 (1997b), 133-49. ------------, Growth Theory in a Keynesian Mode: Some Keynesian Foundations for the Theory of Growth," Journal of Post Keynesian Economics, 19 (Fall 1996a), 113-35. ------------, "Accommodationism Versus Structuralism: Time for an Accommodation," Journal of Post Keynesian Economics 18 (Summer 1996b), 585-94. ------------, “Endogenous Finance,” Chapter 8 in Post Keynesian Economics: Debt, Distribution and the Macro Economy, London: Macmillan, 1996c. ------------, "Debt, Aggregate Demand and the Business Cycle: An Analysis in the Spirit of Minsky and Kaldor," Journal of Post Keynesian Economics, 16 (Spring 1994b), 371-90. ------------, "Competing Views of the Money Supply: Theory and Evidence," Metroeconomica, 45 (February 1994a), 67-88.
32
------------, “Milton Friedman and the Monetarist Counter-Revolution: A Re-appraisal,” Eastern Economic Journal, 19 (Winter 1993), 71 - 81. ------------, “Bank Lending, Discount Window Borrowing, and the Endogenous Money Supply,” Journal of Post Keynesian Economics, 10 (Winter 1987/88), 282 - 301. Pollin, R., "Two Theories of Money Supply Endogeneity: Some Empirical Evidence," Journal of Post Keynesian Economics, 13 (Spring 1991), 366-96. Poole, W., "Optimal Choice of Monetary Policy Instrument in a Simple Stochastic Macro Model," Quarterly Journal of Economics, 84 (1970), 197-216. Robertson, D.H., "Mr. Keynes and the Rate of Interest," in Essays in Monetary Theory, Staples, 1940. Rousseas, S., "The Spheres of Industrial and Financial Circulation Revisited," Economies et Societes, 9 (1994), 315 - 28. Selgin, G.A., and White, L.H., "The Evolution of Free Banking," Economic Inquiry, 25 (1987), 439 - 57. Stiglitz, J., “Capital Markets and Economic Fluctuations in Capitalist Economies,” European Economic Review, 36 (1992), 269 - 306. Stiglitz, J., and Weiss, A., “Credit Rationing in Markets with Imperfect Information,” American Economic Review, 71 (1981), 393 - 410. Tobin, J., Asset Accumulation and Economic Activity, Chicago:Chicago University Press, 1980. ----------, “Keynesian Models of Recession and Depression,” American Economic Review, 1975, 195 - 202. ----------, "Friedman's Theoretical Framework," in R.J.Gordon (ed.), Milton Friedman's Monetary Framework, Chicago:Chicago University Press, 1974. ----------, “Money and Income: Post Hoc, Ergo Propter Hoc?” Quarterly Journal of Economics, 84 (May 1970), 301 - 17. ----------, "Money and Economic Growth," Econometrica, 33 (October 1965), 671-84. -----------, “Commercial Banks as Creators of Money,” in Deane Carson (ed.), Banking and Monetary Studies, Homewood: ILL., 1963. Trautwein, H-M., “The Credit View, Old and New,” Journal of Economic Surveys, 14 (April 2000), 155 - 89.
33
---------------------, “The Uses of the Pure Credit Economy,” in A.Cohen, H.Hagermann, and J. Smithin (eds.), Money, Financial Institutions and Macroeconomics, Boston: Kluwer, 1997. Whitman, M.V.N., “Monetarism and the Monetary Approach to the Balance of Payments,” Brookings Papers on Economic Activity, 3 (1975), 491 - 536. Wicksell, K., Interest and Prices: A Study of the Causes of Regulating the Value of Money (1898), reprinted New York: Kelley, 1965.
34
LM
IS
Interest rate
Output
i*
y*
Figure 1 ISLM representation of classical monetarism.
